Using adaptive models in statistical forecasting of stock prices in Lebanon

Abstract

Nowadays, there are a lot of methods and models to forecast the future values of the stock price, the statistical methods are one of them. The statistical methods have proved to be efficient in the study of time-series, and especially, the exponential smoothing methods have become very popular between researchers due to their robustness. This research explores time-series analysis of three stocks of Beirut Stock Exchange in three different sectors of the economy over the period from 2017 through 2020. It provides analysis based on exponential smoothing methods daily time-series data, deduce the forecasting methods using MAPE and SSE with the best value of α. It was determined that the adaptive forecasting methods (Brown’s model) can be effectively applied for daily forecasting values of Lebanese stock prices, and using historical N days stock price on its own can provide a relatively accurate prediction on N+1 day’s stock price